European Union Antitrust Policy in the Digital Era

The European Commission wants Google to become less connected, the German government wants Facebook to use less data, and the French government wants to promote national champion industries. These policy positions, embodied in recent court rulings, seem to misunderstand the business models underlying the modern tech industry.

Google deals in connections. A search engine that does not allow users to find results on the Internet is useless. Facebook deals in data. Because it has access to data, it can link users and provide its services free of charge. If Facebook were to use less data, it would either be unable to allow its users to connect with one another or it would have to charge them.

These examples illustrate the current state of European antitrust, or competition law. This paper provides a primer on the European Union’s (EU) antitrust regime and discusses how its protectionist and activist bent curbs innovation by adopting an interventionist approach that fails to grasp the economics of the digital era.

European antitrust, or competition, policy is especially salient now that the European Union is pursuing a “modernization” of its antitrust policy and enforcement apparatus. In 2019, the European Commission, the EU’s executive body, published a report commissioned by its Directorate General for Competition on how to modernize antitrust policy. It suggests giving more leeway in doctrine, enforcement, and decision making to EU competition authorities.

There also has been antitrust enforcement activity at the national level. Germany’s antitrust case against Facebook is based on an alleged infringement of European data protection rules under the EU’s General Data Protection Regulation.

There is a benevolent and a less benevolent interpretation for making sense of this state of affairs.

Under a benevolent interpretation, these examples show how little European institutions and governments understand economics. This lack of understanding applies both to linear and to multi-sided business models.

Linear, or traditional, companies create value in the form of goods or services and then sell them to somebody downstream in their supply chain.

Multi-sided businesses, or platforms, are focused on building and facilitating a network. As a specific case of a multi-sided business model, digital, or online, platforms usually do not own the means of production. Rather, they connect the owners of means or producers with those in need of a product or service.

Under a less benevolent view, EU institutions and governments understand the economics of linear and multi-sided businesses, but choose to disregard it in order to advance two political aims—protectionism and consumer welfare (as they conceive the latter). The outcome of this choice is reflected by the EU’s poor scoring in innovation in the digital economy.

There are three important differences between the U.S. and EU antitrust regimes.

  1. In the U.S., the Department of Justice, the Federal Trade Commission, and the states act as plaintiffs, bringing cases to court, to be decided by a judge or panel of judges. In the EU, the European Commission is the sole antitrust agency. It decides which cases to pursue and then decides those cases in first instance.
  1. In the U.S., there is a considerable body of private litigation in antitrust. In the EU, there is not, though that may change. Private litigation occurs mostly at the member state level. However, the EU Damages Directive was intro- duced in 2014 to create an EU law framework for private litigation. In the EU, a directive is a legislative act that sets out certain goals for member states but does not mandate the means to accomplish them.
  2. In the U.S., Section 2 of the Sherman Act deals with monopolization and attempts toward it. Article 102 of the Treaty on the Functioning of the European Union deals with abuse of market power by firms that are already dominant in a given market. While EU law does not prevent a firm from attaining a dominant position in the market, it forbids said firm from pursuing some practices, such as fidelity rebates or predatory pricing. In the U.S., the law focuses on the willful acquisition or maintenance of that power.

Recent European antitrust policy and enforcement are based on a series of economic and legal misunderstandings of how markets work, are driven by an activist agenda, and can have protectionist effects. However, the situation is not entirely dire. In its decisions, the European Court of Justice regularly curtails the Commission’s agenda. In addition, European law offers important lessons on which courts can rely to curb the Commission’s political activism.

Antitrust regulations are applied to all business models alike, taking into consideration the specific economics of each case. Competition law is also business model agnostic. At least in its literal text as well as in the practice of the European Court of Justice, antitrust provisions do not differentiate between online and offline, or between any other business model. It should continue to do so.

Not every competitive advantage results from an anticompetitive practice. Both the text of the law and the Court’s practice routinely accept cooperation between enterprises, and even the market dominance of a single enterprise, if they lead to greater benefit for consumers, more innovation, or a more efficient use of resources.

A practice that makes it more difficult for rivals to compete does not necessarily have anticompetitive effects. European antitrust authorities know that competition is hard, both for incumbent market players as well as for potential entrants.

EU competition law focuses mainly on the exclusion of equally efficient rivals. In its literal text as well as in the Court’s practice, antitrust is not a means to protect inefficient competitors or obsolete sectors and business models.

Anticompetitive effects must be apparent. The Court is unwilling to accept just any claim about potential anticompetitive effects at face value. It demands proof or at least a high plausibility of anticompetitive effects.

A causal link between the anticompetitive practice and the effect must be established. Contrary to the European Commission’s agenda, a narrative of harm does not suffice. To date, the Commission has yet to provide sufficient evidence of anticompetitive behavior in the Google case. At least from an economic point of view, evidence after harm is insufficient to take action.

Markets cannot be judged only by their potential outcomes, but by their inputs and structure, as well.

Introduction

The European Commission wants Google to become less connected, the German government wants Facebook to use less data, and the French government wants to promote national champion industries. These policy positions, embodied in recent court rulings, seem to misunderstand the business models underlying the modern tech industry.

Google deals in connections. A search engine that does not allow users to find results on the Internet is useless. Facebook deals in data. Because it has access to data, it can link users and provide its services free of charge. If Facebook were to use less data, it would either be unable to allow its users to connect with one another or it would have to charge them.

These examples illustrate the current state of European antitrust, or competition law. This paper provides a primer on the European Union’s (EU) antitrust regime and discusses how its protectionist and activist bent curbs innovation by adopting an interventionist approach that fails to grasp the economics of the digital era.

European antitrust, or competition, policy is especially salient now that the European Union is pursuing a “modernization” of its antitrust policy and enforcement apparatus. In 2019, the European Commission, the EU’s executive body, published a report commissioned by its Directorate- General for Competition on how to modernize antitrust policy. It suggests giving more leeway in doctrine, enforcement, and decision making to EU competition authorities.

There also has been antitrust enforcement activity at the national level. Germany’s antitrust case against Facebook is based on an alleged infringement of European data protection rules under the EU’s General Data Protection Regulation.

There is a benevolent and a less benevolent interpretation for making sense of this state of affairs.

Under a benevolent interpretation, these examples show how little European institutions and governments understand economics. This lack of understanding applies both to linear and to multi-sided business models.

Linear, or traditional, companies create value in the form of goods or services and then sell them to somebody downstream in their supply chain.

Multi-sided businesses, or platforms, are focused on building and facilitat- ing a network. As a specific case of a multi-sided business model, digital, or online, platforms usually do not own the means of production; rather, they connect the owners of means or producers with those in need of a product or service.

Under a less benevolent view, EU institutions and governments understand the economics of linear and multi-sided businesses, but choose to disregard it in order to advance two political aims—protectionism and consumer welfare (as they conceive the latter). The outcome of this choice is reflected by the EU’s poor scoring in innovation in the digital economy, something the EU’s own 2019 report, The Digital Economy and Society Index, notes.1

An Overview of EU Antitrust Policy

There are three important differences between the U.S. and EU antitrust regimes.

  1. In the U.S., the Department of Justice (DOJ), the Federal Trade Commission (FTC), and the states act as plaintiffs, bringing cases to court, to be decided by a judge or panel of judges. In the EU, the European Commission is the sole antitrust agency. It decides which cases to pursue and then decides those cases in first instance.
  2. In the U.S., there is a considerable body of private litigation in antitrust; in the EU, there is not, though that may change. Private litigation occurs mostly at the member state level. However, the EU Damages Directive was introduced in 2014 to create an EU law framework for private litigation.2 In the EU, a directive is a legislative act that sets out certain goals for member states but does not mandate the means to accomplish them.
  3. In the U.S., Section 2 of the Sherman Act deals with monopolization and attempts toward it. Article 102 of the Treaty on the Functioning of the European Union (TFEU) deals with abuse of market power by firms that are already dominant in a given market. While EU law does not prevent a firm from attaining a dominant position in the market, it forbids said firm from pursuing some practices, such as fidelity rebates or predatory pricing. In the U.S., by contrast, the law focuses on the willful acquisition or maintenance of that power.3

The sole EU antitrust agency is the European Commission’s Directorate General for Competition, which both develops policy and prosecutes and decides cases. Thus, the Commission acts as both prosecutor and judge of first instance.

The Commission’s decisions can be appealed before the General Court, which functions as a Court of Appeal, and the European Court of Justice, which functions as a Supreme Court. While there is the possibility of private, civil, action in antitrust, such cases are practically nonexistent in the European context. The reasons for this are manifold, though it suffices to say that in Europe, culturally, antitrust is seen as a function of government.

Meanwhile, member states have different national antitrust policies (which will not be covered here). While the competition policies of EU member states are largely harmonized, some countries’ governments impose additional regulations

The Treaty on the Functioning of the European Union has three main competition provisions:

  1. Article 101. Control of collusion and other anti- competitive practices.
  2. Article 102. To prevent the abuse of firms’ dominant market positions.
  3. Article 107. Control of direct and indirect aid given by member states of the European Union to companies.

In addition, the European Union Merger Regulation oversees proposed mergers, acquisitions, and joint ventures involving companies that have a certain, defined amount of sales turnover in the EU.4

The Protectionist Effects of EU Competition Policy

European antitrust policy is solely the responsibility of the European Commission, which defines policies, makes decisions on cases involving those policies, and enforces those decisions. The Commission is a political organ; its members are mostly former cabinet members of the various European Union member states’ governments. They are not elected to the Commission by the public, but appointed in a process defined by member-state and party politics.

Members of the Commission hold on to their party affiliations following their appointments.5

Remarkably, the European Commission’s largest and highest profile cases concerning digital and online businesses have been against American companies (which may be due to the fact that most dominant digital firms are U.S.-based).6 First, it was Microsoft, then Intel, and now Google.7 According to the Commission’s reasoning, there are several problems with these companies. One is their market share. As dominant agents, the Commission believes, they are bound to eventually abuse their market power. The other is their providing goods or services that can be considered essential, which can further strengthen these companies’ dominance over their markets.

In European public discourse, these companies are also resented for being large, U.S.-based multinationals and for being disruptive and threatening to drive their less innovative European competitors out of business. While this sentiment prevails in public discourse, it has not been part of the Commission’s argument. Some Commissioners, however, have expressed similar opinions.8

Consumer welfare is the second political antitrust aim of the Commission and most EU member states. While the consumer welfare standard is widely used by many other competition authorities, including in the U.S., where it is the central standard, the European Commission does not necessarily abide by the standard, but uses the phrase consumer welfare often as a political goal and independent from the widely used antitrust standard.9 This gives EU authorities considerable power to act and combine enforcement of competition policy with political aims.10

This is particularly problematic because businesses subject to antitrust action have to show that their conduct does not affect consumers negatively.11 Thus, in the Commission’s understanding of it, the EU’s consumer welfare aim effectively turns the burden of proof on its head by requiring companies to prove their innocence.12 However, it is important to note that European courts have been less willing to accept the leeway implied here. In recent years, courts have been striving for an interpretation of consumer welfare more in line with its general use as an antitrust standard and less as a political tool.13

Modernizing EU Antitrust

The European Commission and several EU member states have decided that  it is time for a “modernization” of antitrust policy. Pressures influencing this decision include the emergence of digital business models, the continuous evolution of antitrust policy, differences with the U.S. approach, the recent setting up of more powerful antitrust agencies in Asia and Latin America, and pressure exercised via the International Competition Network, a global network that seeks to facilitate cooperation between competition authorities and the harmonization of their practices.

Given the peculiar structure of the EU, its antitrust regime can be modernized without changing the law. Rather, it is a matter of either adjudication or administrative review. Individuals and businesses can only challenge changes in policy by going to court. Only those organizations involved in antitrust cases can appeal decisions based on these policies in courts of higher instance.

To achieve consensus that a policy change is necessary, the European Commission usually publishes expert views after consulting with stakeholders, such as, for example, academics, consumer-protection groups, and industry groups. A paper is drafted based on this feedback. If the Commission agrees on its contents, the paper becomes a Guideline— official policy that can lead to enforcement and decisions. However, the extent of the changes that can be introduced via guidelines is limited.

Major overhauls require new legislation.

In 2019, the Commission published the report it commissioned, “Competition policy for the digital era,” authored  by Jacques Crémer, professor of economics at the Toulouse School of Economics; Yves-Alexandre de Montjoye, professor of data science at Imperial College in London; and Heike Schweitzer, professor of law at Humboldt University in Berlin. In the report, the authors make several sug- gestions on how to modernize EU antitrust policy in order for it to better deal with digital and online businesses.14

Essentially, the authors advocate making it easier for the European Commission to reverse the burden of proof in cases involving digital and multi-sided business models. The premises underlying this proposal are that the effects on digitization on competition are uncertain, and that the “stickiness” of market power is too great to rely on existing principles and approaches. These premises are formulated after a discussion of three aspects the authors consider crucial to digitization:

  1. Extreme returns to scale;
  2. Network externalities; and
  3. The role of data.15

The report fundamentally misunder- stands these aspects.

Economic Misunderstandings The report claims that there are “extreme” returns to scale in digital markets but does not explain why these returns to scale are more extreme than those that exist, for instance, in network industries like telecommunications or energy that are often seen to have natural monopoly characteristics. They also fail to point out how returns to scale are more “extreme” in digital than in non- digital markets.

Similarly, the report discusses at length network effects in platforms as if they were new. However, there is a considerable body of EU case law dealing with online and offline platforms as well as with network effects. It is European Court of Justice practice not to assume detriment to competition by platforms and network effects. Contrary to the report’s prescription, the Court currently assesses detrimental effects on a case-by-case basis.

This is because there are many cases in which consumers benefit from network effects. For example, consumers benefit from large credit card net- works that are accepted by a wide variety of merchants.16 Another is that users benefit from the technological integration of different products, such as Voice over Internet Protocol communications complementing a computer operating system. An important case is the integration of Skype in the Microsoft Office suite. The Court found that the benefits to users considerably outweigh any potential accumulation of market power.17

In addition to EU case law, there are numerous examples of innovation in a free market environment leading to the demise of once mighty dominant players, be they digital, online, or plat- forms. Major examples include the de- cline of Yahoo! and CompuServe as gateways to the Internet, the shift to- ward the mobile Internet accessed via smartphones rather than computers, and the digital transformation of the media landscape.18

The report stresses that data is important, but it does not say why and how it is important in regard to antitrust. On the contrary, it states:

The significance of data and data access for competition will always depend on an analysis of the specificities of a given market, the type of data, and data usage in a given case.19

The report then goes on to say that “any discussion on (access to) data must take into account the heterogeneity of data and its uses along many dimensions.”20 How can data, if so case-specific and heterogeneous, act as an indicator for the need of more robust enforcement or for a new policy?

Worse, the authors do not provide empirical evidence for any of their claims. Instead, they refer to well- known cases like those concerning Google and Facebook. Hence, they commit a twofold mistake. First, they assume that all digital businesses are multi-sided markets, or platforms.

Second, they tacitly assume that platforms tend toward establishing a dominant position in their respective markets.

Naturally, antitrust is only interested in dominant firms, but the fact that platforms are a major focus of the European Commission’s report suggests that the report considers them to be dominant as such. Moreover, the report’s authors do not state specific criteria for determining whether platforms are dominant or not. This becomes clear in their conclusions:

Because of the innovative and dynamic nature of the digital world, and because its economics are not yet completely understood, it is extremely difficult to estimate consumer welfare effects of specific practices. Given the concentration tendencies of platforms, and the high barriers to entry in some of the markets they dominate, a finding that they restrict the ability of other firms to compete either on the platform or for the market in a way which is not clearly competition on the merits should trigger a rebuttable presumption of anti-competitiveness. It should be the dominant platform’s responsibility to show that the practice at stake brings sufficient compensatory efficiency gains.

Given the breadth of the presumption, and the fact that our insights into possible countervailing efficiencies are still evolving, such efficiency defenses should be fully explored by competition agencies and courts. [Emphasis added]21

Note how, by this reasoning, innovation and the dynamic nature of the digital world are explicitly recognized, but also turned into an argument in favor of regulatory activism. Simply put, the logic this argument follows is: If something is new, we cannot know its effects. If we cannot know its effects, we must regulate it. And the burden of proof lies solely with the regulated party, not the regulators.

The report also concludes:

Platforms act as regulators of the interactions they host.

If dominant, they have a responsibility to ensure that they regulate in a pro-competitive way.

… Dominant platforms should be subject to a duty to ensure interoperability with suppliers of complementary services.22

Here we see a serious misunderstanding of the economics of exchange. In a contract, the parties to the contract agree to regulate their behavior in order to fulfill the terms of the contract. In this way, a contract acts as a form of private regulation. In multi-sided markets, as in any exchange, there is a natural regulation of the exchange by the parties to the exchange. In the case of the platform, it is the platform itself, as well as the parties to the markets that platforms help to intermediate. The same is true for a kid running a lemonade stand—sellers and buyers regulate the conduct of their exchange.

Against this background of misunderstandings, the authors of the EU report make several suggestions on how to modernize European antitrust. These are:

  1. “[E]ven where consumer harm cannot be precisely measured, strategies employed by dominant platforms aimed at reducing the competitive pressure they face should be forbidden in the absence of clearly documented consumer welfare gains.”23
  2. “[L]ess emphasis on analysis of market definition, and more emphasis on theories of harm and identification of anti- competitive strategies.”24
  3. “[E]rr on the side of disallowing potentially anticompetitive conducts, and impose on the incumbent the burden of proof for showing the pro-competitiveness of its conduct.”25
  4. “[C]ompetition law enforcement and regulation are not necessarily substitutes, but most often complements and can reinforce each other.”26

These suggestions are interventionist measures that disregard many aspects of due process.

Suggestion a) translates into a doctrine of “when in doubt, ban it.” It is impossible to reconcile this with economic freedom and innovation. It is difficult to picture how a business could irrefutably prove its consumer welfare case. The political use of consumer welfare—as generally practiced by the Commission and specifically proposed in the report— places the burden of proof entirely on the company. For instance, it is difficult to imagine rideshare companies like Uber and Lyft ever getting off the ground had they first tried to get the regulatory go-ahead from local taxicab commissions.

Suggestion b) is accessory to a) in making it easier for the European Commission to bring antitrust enforcement action against any business model it chooses. This is especially important since the report claims that the market dominance of digital businesses is a problem.

Dominance can only occur in a clearly defined market. But suggestion b) claims the opposite. In essence, it suggests that simply saying that a digital business is dominant makes it so. The market need not to be clearly defined. Instead, the Commission can construct a narrative of harm. From that narrative, it automatically follows that the company is dominant.

Suggestion c) reinforces the inversion of the burden of proof contained in a) and b).

Finally, suggestion d) hints that antitrust is not sufficient to address challenges of the digital age, and therefore sector-specific regulation addressing digital business models is needed. The EU is considering this, as well.27

In short, the report seems to suggest that the Commission’s approach to adapting competition policy to the digital era is to not allow only a narrow set of digital businesses and business models to emerge, after they have cleared all the competition policy hurdles.28 And those that do make it through this process are unlikely to be among the most innovative. Thus, it is no surprise that some incumbents support the ideas expressed in the report. For example, the German Federation of Industries, known by its German initials BDI (Bundesverband der Deutschen Industrie) fully supports the suggestions.29

Another, perhaps equally important problem with the report’s suggestions is their lack of definition and differentiation regarding what constitutes a digital business. As noted, it is unclear what the authors consider a digital business. They use the term synonymously with multi-sided markets, or platforms. While the authors offer a list of characteristics of these business models, they do not explain if the list is exhaustive and do not state the criteria for including them in the list. Yet, whatever the reasoning behind the list, it remains unclear as to what type of platform it refers—and there are several different types of platform. It also remains unclear if incumbents in the process of digitizing their business models also fall under the “modern” antitrust proposed in the report.

Siemens, for example, tracks household electricity consumption.30 This data could be used to improve the technology of pumps and turbines, to develop new applications, and to sell it on aggregated level. So, Siemens is using data. Does it follow that it is a digital business? Does it follow that it must prove beyond doubt that all consumers always benefit from this data collection? Danone is digitizing its business model through customer interaction. Do the suggestions apply here, too? Or consider a retailer, such as Walmart, that invests heavily in building its online operations. At what point does it become a digital business? All of these examples involve “extreme” effects of scale. Does it follow that they are to be treated like digital businesses? As these simple examples show, the approach endorsed by the report is unclear and poses more problems than it aims to solve.

Another problem is the amount of leeway the suggestions leave to the European Commission. In effect, the suggestions say that whatever the Commission wants is automatically true, unless it is disproved.

These suggestions undermine the presumption of innocence that is central to modern due process. Today, if the Commission wants to prosecute an antitrust case, the burden of proof normally rests with the Commission. With the adoption of the report’s suggestions, this burden would be inverted in some cases. The Commission would formulate a narrative of harm and the prosecuted agents would have to disprove it.

Moreover, the Commission likely will get some important cases wrong. For instance, in May 2020, the European General Court annulled the 2016 decision by then-EU Competition Commissioner Margarethe Vestager to block the merger of the mobile phone companies O2, owned by Spain’s Telefonica, and Three, owned by Hong Kong’s Three. The Court ruled that the Commission had made “several errors of law” in calculating the potential harm from the merger, and said it had not proven that deal would lead to higher prices or reduced competition as a result.31

Another problem—compounding the problems from the lack of differentiation—is that the report is ambiguous on whether this suggestion should apply only to digital markets or to all markets. Since the report focuses on digital markets, it appears that the reversion of the burden of proof is only suggested for digital markets.

However, this would mean a bifurcation of EU antitrust law. On the one hand, defendants in offline cases would still enjoy a presumption of innocence, while in online cases the presumption would be based on the Commission’s narrative of harm, which has to be refuted by the defendant.

This is not only unfair, but impracticable. In an economy witnessing the continuous merging of online and offline business models, developing two different types of antitrust practices is not feasible. Does that mean that the Commission’s report envisages reversing the burden of proof in all cases? The report does not address this question.

Ultimately, the report represents the view of the European Commission. While written by independent authors, it was shaped by continuing interaction with the Commission and pressure groups. The Commission welcomed the report and is in the process of preparing new Guidelines based on it. What makes the report both dangerous for innovation and appealing to the Commission is the great amount of leeway the Commission would gain in both enforcement and decision mak- ing, if the report’s policies were to be adopted.

But not all is bad news. In any event, the changes suggested—whether introduced on a case-by-case basis or via new sets of Guidelines—cannot be implemented by the Commission alone, since they would have to be validated, ultimately, by the European Court of Justice. That is not a foregone outcome. In fact, on issues of proof, evidence, and effects (let alone the issue of expert consensus mentioned above), case law has markedly moved in the opposite direction. Would the EU courts change course now? That remains to be seen.

To illustrate the degree of economic and legal misunderstanding in the European Commission’s actions, the following section presents two case studies—the Commission’s Google Shopping case and the German Facebook case. Both case studies show that the suggestions made in the report are already the reality of antitrust policy in the EU. The case studies also show how antitrust enforcement can help advance an activist and protectionist agenda.

Case Study 1: The European Commission’s Google Shopping Case

The European Commission holds the view that Google has systematically given prominent placement to its own comparison-shopping service while demoting competitors in its search results.32 The Commission holds that this practice amounts to an abuse of Google’s dominant position in general online search by stifling competition in comparison shopping markets.

In this instance, the Commission’s case serves as a good example of a persistent problem in antitrust law in general: definition of the relevant market. It is always problematic because it inherently involves judgement calls that involve some arbitrary drawing of lines around said market.

That applies in this case. One of the decision’s main problems is that it is unclear as to whether there is a market for general Internet search and, if so, who is active in that market and whether Google Shopping is part of a separate comparison-shopping market. According to the Commission, consumers go to Google Shopping— or to similar European online shopping platforms such as Foundem or idealo.de—just to compare shopping opportunities and not primarily to buy. On the other hand, the person who browses through Amazon, eBay, or Zalando—a German online retailer specializing in apparel—are primarily motivated to buy.

Thus, according to the Commission, a person browsing through a platform has no intention to buy, while a user browsing an online marketplace has a well-formed intent to buy. However, the Commission does not specify how the distinction between a platform and a marketplace is to be drawn. It also does not offer empirical evidence in favor of its finding.

This assumption is not based on any study or evidence of consumer behavior, but merely on the fact that Google Shopping does not offer products for sale, only links to other retail websites. It also implies that the market definition would change should Google Shopping introduce one-click shopping, vertically integrate into retailing products itself, or develop its own marketplace. Ignoring Amazon, eBay, Zalando, and the like as competitors for Google Shopping without studying actual consumer behavior appears negligent when defining an online shopping market, but it fits well within the European Commission’s broad narrative of harm.

In addition, Google Shopping does not portray itself as a neutral general search engine. The Commission’s analysis simply ignores the fact that Google Shopping is clearly labeled as an advertising service, with markers or notices telling users that Google is paid for these ads.33 Hence, it is not clear how many consumers expect a “neutral” listing of Google Shopping results.

The Commission is using double standards regarding Google. It maintains that Google is abusing its market power by privileging its own comparison-shopping site. But why, exactly? The Commission seems to believe that the link is that Google Shopping is part of Google, but other platforms and marketplaces—to use the Commission’s own distinction— have similar practices and have never been considered abusive under Article 102 of the TFEU.

In response to the Commission’s decision, Angela Daly, law professor at the University of Macerata in Italy, asks, what is “[t]he extent to which Google’s conduct constitutes behavior previously recognized to comprise an infringement of EU competition law, since it does not fall squarely into a recognized ‘head’ of abuse?”34 And Jakob Kucharczyk, an attorney and vice president of the European Computer and Communications Industry Association, in a 2017 paper, states:

[C]ontroversies are found at every step of the Commission’s analysis: the definition of the relevant market, establishment of dominance, and the finding of abuse. … However, the Commission wrongly seems to view Google as something like the “master of the [I]nternet,” deciding single-handedly over the fate of other online companies.

But strangely, it seems that the vast majority of successful online businesses decided to take their fate into their own hands.35

In short, the European Commission has decided to politicize market definition regarding online business models, specifically in the investigation launched against Amazon on July 17, 2019.36 By doing so, it is applying some of the suggestions discussed in the report to advance an activist agenda. It remains to be seen how the courts rule in appeals of the case.

Case Study 2: The German Facebook Case

In 2019, Germany’s Federal Cartel Office (FCO, Bundeskartellamt) prohibited Facebook from combining user data from different sources.37 Although this case is not being prosecuted by European Union authorities, it is an interesting case study. It illustrates how the suggestions made in the European Commission’s report are already the reality of antitrust enforcement within EU member states. Moreover, the German Facebook decision appears to have strongly influenced the Commission.38

At the beginning of its investigation, the FCO put out the following narrative of harm: Since many platforms do not charge both sides of their market, but only one of them—advertisers, but not users—then new forms of exploitative abuse may emerge, such as demanding too much data.

The FCO investigated Facebook’s behavior vis-à-vis its users. It found that Facebook has a dominant position in the market for social networks and that Facebook’s general terms and conditions are inadequate and therefore constitute an exploitative abuse of market power. According to Facebook’s terms and conditions, users access the social network under the precondition that Facebook may collect their data outside of the Facebook website or app—on third-party smartphone apps, Facebook-owned services such as WhatsApp and Instagram, and third- party websites—and combine and assign that data to the user’s Facebook account.

In its press release, the FCO states:

In the authority’s assessment, Facebook’s conduct represents above all a so-called exploitative abuse. Dominant companies may not use exploitative practices to the detriment of the opposite side of the market, i.e. in this case the consumers who use Facebook.

This applies above all if the exploitative practice also impedes competitors that are not able to amass such a treasure trove of data.39

Note that the case is entirely based on an alleged infringement of European

Union data protection rules under the General Data Protection Regulation. It is remarkable for a competition authority to determine if a breach of data protection rules has occurred, given that specialized agencies are tasked with examining this issue.

According to the FCO’s reasoning, Facebook’s breach of European data protection rules is an indication of its holding and abusing its dominant position in the market. This is an example of an inverted burden of proof. It also greatly expands the scope of competition policy to cover other matters beside competition, in this case privacy.

Note the absence of an economic theory of harm. Instead, the FCO constructs a narrative in which data serves as a proxy for utility and disutility. The papers published by the FCO to date leave several questions unanswered: How are users harmed if they give up data in order to get services in exchange? Is it true that users give up “too much data”? What harm do they face by doing so? Data being non-rivalrous, how is it that customers, in the FCO’s wording, “give up data”?40

It might be possible to imagine data as a proxy for money or bartered goods. In assuming so, it could be possible to construct an economic theory of harm. Note, however, the faults in this construction. First, the non-rivalrous nature of data makes it difficult to use as a proxy for money. You do not give up data when you share it. Second, according to German law, data cannot be used as money, owned, or transferred as property.41 So, independently of the merits of monetizing data, by treating data as a proxy for money, the FCO goes beyond German law.

What about the remedy recommended by the FCO? Facebook can continue to operate only if it collects less data. In the absence of an economic theory of harm, this remedy undermines economic freedom and can curb competition. Social network users are not harmed by sharing personal data and having data sets combined.

Consumers cannot be exploited if they do not mind providing the data collected.

If data were as important in leading to economies of scale as the European Commission maintains, then customers benefit from sharing it.

After all, they use Facebook and other applications free of charge. The amount and the interconnection of data allows Facebook to develop new products based on users’ needs, which, in the absence of a price mechanism, Facebook can only know through data. Less data means fewer products. At the very least, less data leads to fewer interconnected products—and inter- connectivity is what users want from social media.

Additionally, combining data facilitates the development of better technologies to rank news and other information to match users’ interests. Therefore, prohibiting the practice would lead to a deterioration of the services offered, as the matching technology would deteriorate.

At the same time, Facebook would become less competitive in advertising markets vis-à-vis Google and other market participants. Given that data is used to develop and offer better services, preventing Facebook from collecting, combining, and using data would hinder Facebook’s ability to innovate and offer better products and services; both outcomes would harm competition.

In contrast, Facebook users and advertisers benefit from the use and combination of “on-Facebook” and “off-Facebook” data from different sources, which facilitates the improvement of matching algorithms to rank information and news for users. In addition, it is difficult to conceive how users can be exploited by online platforms using their data, given that their data resources are not depleted when used and that most people tend to willingly share data in order to obtain benefits such as improved services.

Like the European Commission, Germany’s Federal Cartel Office has adopted a theory of harm that elevates a vague notion of market dominance above real-world consumer welfare.

Conclusion

Recent European antitrust policy and enforcement are based on a series of economic and legal misunderstandings of how markets work, are driven by an activist agenda, and can have protectionist effects. The Commission’s views on platforms, the Google case, and Germany’s Facebook decision are examples of this. However, the situation is not entirely dire. In its decisions, the European Court of Justice regularly curtails the Commission’s agenda. In addition, European law offers the following important lessons on which courts can rely to curb the Commission’s political activism.

Antitrust regulations are applied to all business models alike, taking into consideration the specific economics of each case. Competition law is also business model agnostic. At least in its literal text as well as in the practice of the European Court of Justice, antitrust provisions do not differentiate between online and offline, or between any other business model. It should continue to do so. To adopt the recommendation in the European Commission’s report would undermine this and create considerable confusion in the market.

Not every competitive advantage results from an anticompetitive practice.42  Both the text of the law and the Court’s practice routinely accept cooperation between enterprises, and even the market dominance of a single enterprise, if they lead to greater benefit for consumers, more innovation, or a more efficient use of resources.

A practice that makes it more difficult for rivals to compete does not necessarily have anticompetitive effects.43 European antitrust authorities know that competition is hard, both for incumbent market players as well as for potential entrants.

EU competition law focuses mainly on the exclusion of equally efficient rivals.44 In its literal text as well as in the Court’s practice, antitrust is not a means to protect inefficient competitors or obsolete sectors and business models.

Anticompetitive effects must be apparent.45 The Court is unwilling to accept just any claim about potential anticompetitive effects at face value. It demands proof or at least a high plausibility of anticompetitive effects.

A causal link between the anticompetitive practice and the effect must be established.46 Contrary to the European Commission’s agenda, a narrative of harm does not suffice.

To date, the Commission has yet to provide sufficient evidence of anticompetitive behavior in the Google case. At least from an economic point of view, evidence after harm is insufficient to take action.47 Markets cannot be judged only by their potential outcomes, but by their inputs and structure, as well.